Monday, January 17, 2011

Theropods That Fit the Bill

[Image: Skull of a gull compared to a living specimen. The apparent differences in the extent of the beak are caused by the presence of different kinds of rhamphotheca.]

One aspect of Mesozoic birds and other theropods that doesn't get discussed much are their bills. Beaks or beak-like structures (generally, rhamphotheca) have a wide but extremely spotty distribution among Mesozoic theropods, and numerous theropod groups seem to have evolved them independently of one another.

Perhaps the most basal known theropod that likely bore a beak (assuming we can safely exclude the skull material attributed to the infamous "first bird" Protoavis) is Limusaurus inextricabilis from the Oxfordian, about 160 million years ago. Though it belonged to the otherwise toothy clade Ceratosauria, Limusaurus was completely toothless. To be precise, most theropods have teeth in the three major sections of their jaws: the premaxilla at the front of the skull, the maxilla behind it, and the dentary, or lower jaw. Ornithischian dinosaurs had an additional bone, the predentary, at the tip of the lower jaw that was always toothless, indicating that almost all ornithischians probably bore a beak, at least on the lower jaw (and ceratopsians added an additional, pre-premaxilla: the rostral). Limusaurus lacked teeth in the premaxilla, maxilla, and dentary, so it was completely toothless and, therefore, it is assumed to have been beaked.

[Image: Skull reconstruction of Limusaurus, posted by Dave Hone at his Archosaur Musings blog.]

As far as I know, Limusaurus is the only non-coelurosaurian theropod that bore a beak. Generally, this fits well with the idea that non-coelurosaurian theropods were hypercarnivorous flesh eaters with little variation in their diets. However, as a number of recent papers by Lindsay Zanno and colleagues has shown, coelurosaurs started playing around with omnivory and herbivory, and as a consequence, many lineages cast off some or all the sharp teeth that characterized their ancestors.

[Image: Skull reconstruction of Pelecanimimus by Ville Sinkkonen]

The first coelurosaur lineage to start shedding their teeth permanently were the ornithomimosaurs. This group is well-known for their toothlessness, and evidence from specimens with preserved keratin sheaths on the jaws (Norell et al, 2001) show definitively that they bore keratinous beaks on both the upper and lower jaws, at least in advanced species. As expected, the most primitive ornithomimosaurs still retained teeth. Pelecanimimus polydon, as its name implies, had teeth in the premaxilla, maxilla and dentary. The next most advanced ornithomimosaur, Harpymimus, is the first example among theropods of the "half beak." While its upper jaw was toothless and likely beaked, it retained teeth in the lower jaw. I'll come back to the implications of these "half beaks" further down. The most basal ornithomimosaur wich lacked teeth completely seems to be Garudimimus, though the status of Beishanlong is unknown.

After the ornithomimosaurs, beaks are found in therizinosaurs. Again, early forms like Falcarius had a full compliment of teeth, but in laker ones like Beipiaosaurus, the teeth became restricted to the middle of the jaws, leaving room for a beak in the front, similar to ornithischians. Next, oviraptorosaurs seem to have developed beaks fairly early in their evolution. While Incisivosaurus had teeth in both jaws (though restricted to the tips), Caudipteryx had only a few teeth in the premaxilla, another example of a "half beak" where a toothless lower jaw is matched with a toothy upper jaw. While the several possible intermediate forms lack skulls, both caenagnathids and oviraptorids proper had fully toothless jaws, so presumably they were fully beaked by the time they diverged.

Next up were the deinonychosaurs, and while some seem to have dabbled in omnivory, they were mainly carnivorous, and none seem to have developed beaks. Greg Paul is infamous for restoring them with cornified tissue or "proto beaks" at the tips of their jaws adjacent to teeth, but there isn't much evidence to support this idea as far as I know. However, many deinonychosaurs which preserve feathers show a (usually small) portion near the tip of the snout that is unfeathered. This featherless snout tip is also seen in some toothed, presumably beakless enantiornithines. It's possible this could be evidence of "rhamphotheca" in its loosest sense--the very lightly cornifies, flexible bill skin found toward the back of the beaks in some modern birds, where the horn-like, kerationous portion thins out into normal skin. More on these flexi-bills down the page.

[Image: My own simplified cladogram of Aviremigia showing some notable occurrences of beaked and half-beaked and beak-free birds. Excuse the quality in close-up, some of these were yoinked from works in progress.]

Around the base of Avialae, the occurrence of half-beaks seems to explode. Almost all known basal, non-pygostylian avialans lack teeth in the upper jaw (e.g. Jeholornis, Yandangornis) or lower jaw (omnivoropterygids) and, possibly, both (Jixiangornis preserves no teeth and may have been full-beaked, but the skull is badly crushed, and crushing has been known to obscure the tiny teeth that should be present in other basal avialan specimens). This trend appears to culminate with the confuciusornithids, which are not only toothless but have sharply pointed jaws that, in some very rare specimens, preserve the actual keratin of the beak. These impressions show that even in these fully beaked birds, the rhamphotheca was thin and delicate and probably not as heavily mineralized as in modern birds.

The next bit is rather odd given the trends seen at the base of avialae. In most basal ornithothoraces (the group that includes enantiornithines and ornithuromorphs), the jaws are fully toothed, with no evidence for beaks. It's tempting to think that this could unite the enantiornithines with the toothy, beakless deinonychosaurs and Archaeopteryx in a "Sauriurae" to the exclusion of the beaked birds. However, given the numerous times beaks have evolved independently in vertebrates, it's not unthinkable that each of the half-beak examples above arose independently of one another (or, that some reversal occurred at the base of ornithothoraces to return birds to a state of fully-toothed maws). Either way, while many enantiornithines preserve jaw material, only a few exhibit the kind of toothlessness at the front of the jaws that could imply a beak. These include the half-beaked Alethoalaornis (with a toothless dentary but toothed premaxilla), the fully-beaked Gobipteryx, and the half-beaked Boluochia. Boluochia had teeth in the dentary but none in the upper jaw and, in fat, had a strongly hooked premaxilla similar to the hooked beaks of modern raptors. Enantiornithine relationships are too unresolved to be able to tell if these beaked birds all form a natural group, or if beaks evolved multiple times among enantiornithes, but given their diversity it's a safe bet that many different lineages came up with their own particular feeding solutions.

[Image: The compound rhamphotheca of an albatross is visible as distinct sutures, and is also found in some non-avian birds. Image from the RSPB.]

As expected based on the enantiornithine record, most primitive ornithuromorphs (the fan-tailed birds, including Aves) either lakced beaks or exhibited a half-beaked configuration. While Hongshanornis was originally reported to have a beak, Jingmai O'Connor and colleagues later showed that it had tooth sockets preserved in the upper and possibly lower jaws (same goes for the related Longicrusavis). The earliest fully-beaked ornithuromorph is also one of the most primitive, however: Archaeorhynchus lacked teeth and had a flattened, Spoonbill-like beak. The songlingornithids (such as Yixianornis), and the later hesperornithines and Ichthyornis all had a therizinosaur-like configuration, with toothless premaxilla (and even predentary-like bones in the hesperornithines) with toothy maxilla and dentaries. Evidence from bone texture shows that likely had keratinous beaks at the tips of their jaws, and either feathery toothed jaws or pliable, skin-like rhamphotheca further back. Since both major lineages of avians lack teeth, it's likely their common ancestor was also fully beaked, so teeth must have been lost for good in the bird lineage shortly after Ichthyornis diverged. Interestingly, studies of ichthyornithines and hesperornithine bone structure shows that they likely had "compound rhamphotheca", and this may have been the ancestral condition for modern birds (Heironymous & Witmer, 2010). While the quintessential bird beak is made up of a single keratin sheet covering the jaw, in species with compound beaks, the keratin is arranged in discrete plates on the jaws. This can best be seen in some modern seabirds like the Albatross.

Unfortunately, in interesting taxa like Hollanda, Gansus, and Patagopteryx, the condition is unknown. However, we can use parsimony and phylogenetic bracketing to try and come to a reasonable guess. Most studies find these three to be ornithuromorphs more primitive than the hesperornithines and Ichthyornis. Given the condition in Ichthyornis, it's most likely that Archaeorhynchus evolved its toothless beak independently of more advanced birds, most of which have toothed upper and lower jaws with beaked tips in the grade between Archaeorhynchus and Aves. While the three intermediate birds may well have lost some or all of their teeth independently again, it's probably more parsimonious to assume that like hesperornithids, hongshanornithids and songlingornithids, they were either fully toothed or had beaks restricted to the tips of the jaws.

[Image: Cassowary chick by Michael Thirnbeck, from Flickr. Note the horny beak only covers the tip of the bill, leaving the nostril surrounded by more skin-like tissue. Thanks to meidamon over at DinoForum for finding this and other pictures illustrating beak anatomy.]

So, to rephrase an old fallacy, what use is half a beak? Many Mesozoic coelurosaurs had teeth only on the upper or lower jaw, and most researchers have assumed that edentulous = beaked (see for example Zanno et al., 2009). But "beaks" are more complex than simple sheathes of keratin covering toothless parts of the jaw. As anatomist Larry Witmer has explained (in print as well as on the DML), rhamphotheca can consist of everything from solid, hard keratin (which, obviously, is a bit antithetical to tooth growth and replacement) to softer, pliable, barely mineralized tissue. For example, it's a general rule of thumb that the keratin rhamphotheca never completely encloses the external nares, meaning the nostrils of a bird are never encased in keratin. This is most obvious in species that have a cere, a prominent fleshy part of the posterior rhamphotheca found in pigeons, parrots and hawks, among other birds. But in many other birds (especially aquatic forms like ducks and gulls), the nostrils do appear to erupt from mid-bill. What's going on?

[Image: Duck skull by ~Oryx gazella~, from Flickr. Note that only the tip is fully keratinized.]

Basically, the "beak" of a duck (and many other birds) is only really solid towards the tip. The broad bill of a duck is, laterally, fairly soft, and made of rhamphotheca that is only lightly mineralized. Essentially, it's mostly toughened skin (this is why ducks and other anseriforms are particularly prone to bill injuries). In the kiwi, the nostrils don't just erupt from the "beak", they occur at the very tip. Again, the bill of a kiwi is not highly keratinized, but rather made up of soft, pliable integument full of sensory organs (Martin et al., 2007).

So, our mysterious half-beaked birds may have had some cere-like tissue covering the upper portion of the toothed bill, while the lower portion contained more keratinized rhamphotheca. Perhaps the upper jaw had a small keratinized tip wedged between the front teeth, as seen in toothed pterosaurs like Rhamphorhynchus and Pterodactylus. Or maybe the half-beaks of some of these birds contained little to no keratin at all. As the kiwi shows, it's not always necessary for all food-aquisition behaviors to have a beak or teeth.

Hopefully, this rough guide can also serve as a reference for artists, many of whom are a bit over-eager to add a beak to anything that is called a "bird." While a wide range of rhamphotheca like structures may have been present on Mesozoic birds, it wasn't necessarily present on all of them, and many probably looked a bit strange compared to the simple, smooth yellow bills of modern birds.

* Heironymous, L. and Witmer, L.M. (2010). "Homology and Evolution of Avian Compound Rhamphothecae." The Auk, 127(3): 590-604. doi: 10.1525/auk.2010.09122
doi: 10.1525/auk.2010.09122
* Martin, G.R., Wilson, K.-J., Wild, J.M., Parsons, S., Kubke, M.F., et al. (2007). "Kiwi Forego Vision in the Guidance of Their Nocturnal Activities." PLoS ONE, 2(2): e198. doi: 10.1371/journal.pone.0000198
* Zanno, L.E., Gillette, D.D., Albright, L.B., and Titus, A.L. (2009). "A new North American therizinosaurid and the role of herbivory in 'predatory' dinosaur evolution." Proceedings of the Royal Society B, 276(1672): 3505–3511. doi: 10.1098/rspb.2009.1029.


  1. Thanks for another handy guide! Everyone knows that ornithomimids, oviraptorids, confuciusornithinds, and neornithines have beaks, but it's hard to find information on all those half beaks.

  2. Thanks very much for this!
    Are there any tell-tale signs on the bone surface that favor or disfavor interpretation of a beak? Of degree of mineralization? Smoothness or pores?

  3. Thanks guys!
    @davidmaas: Yes, there are some distinct things that are usually used to say that a keratious beak was present, though I don't know of anything that would indicate the level of keratinization (or other odd situations like ducks; I didn't get into this in the post, but the pterior bill is not only lightly keratinized, it is actually anchored to a layer of normal unmineralized epidermis, rather than growing directly on the bone).

    Several correlates between horny sheathes (including beaks) and the bone they grew from. Witmer listed the following:
    1. Tangentially oriented rugosity, 2. Dense neurovascular grooves, 3. Oblique neurovascular foramina, 4. Lip at transition to softer skin.

    Clarke et al. 2006 noted that, for example, the forward toothless portion of the premaxilla and dentary in Yixianornis is especially pocked with foramina corresponding to the lack of teeth, implying that the tip was beaked. This type of bone texture has also been used to infer beaks in Ichthyornis, therizinosaurs, and ornithischians, though most papers don't really describe the grooves and foramina in enough detail to suss out if it's really a beaked area or vascularization for some other kind of soft tissue (see the whole theropod lip debate). It seems pretty reasonable to me to assume these represent beaks when they correspond with toothless regions of the jaws.

    Both the Witmer and Clarke papers are free online: (pdf) for an in-depth discussion of this sort of thing.

  4. Thanks for this! Need to do some comparative sketches... and find some skulls.